role in andclinical aicardi - british journal of ophthalmology · aicardi syndrome seen at a local...

British Journal of Ophthalmology 1996;80:805-81 1

Role of ocular involvement in the prediction ofvisual development and clinical prognosis inAicardi syndrome

Allison V Menezes, Terri L Lewis, J Raymond Buncic

AbstractAimsibackground-This study was un-dertaken to document visual function andacuity in patients with Aicardi syndrome,and to determine whether there is anyrelation between ocular features of thesyndrome exhibited at birth and latervisual function.Methods-Fourteen patients with Aicardisyndrome, all examined and followed bythe same ophthalmologist, were reviewedbetween 1975 and 1992 and their ocularcharacteristics and visual acuity de-scribed. It was hypothesised that largerlacunae may be associated with poorerclinical outcome and therefore the rela-tion between these two variables wasinvestigated.Results-Visual acuity as documented bySnellen, Sheridan-Gardner, preferentiallooking, or pattern visual evoked potentialtests was in the normal to low normalrange in six eyes of four patients. Visualfunction correlated significantly withmacular appearance. Good visual func-tion was preserved if the fovea appearednormal on funduscopic examination andwas uninvolved by lacunae. The size ofthelargest chorioretinal lacuna also corre-lated significantly with clinical outcome:patients with large lacunae were morelikely to be immobile and to have nolanguage skills.Conclusion-It was concluded that goodvisual function in patients with Aicardisyndrome may be anticipated if the foveais normal. Although many patients havesevere psychomotor retardation, the pres-ence ofpredominantly small chorioretinallacunae may indicate a better prognosisfor mobility and language development.(BrJ Ophthalmol 1996;80:805-81 1)

Aicardi syndrome, first described in 1965,' ischaracterised by the clinical triad of infantileseizures, agenesis of the corpus callosum, andchorioretinal lacunae. Most authors considerthe lacunae to be pathognomonic of the disor-der.2 To date, all undisputed cases of thesyndrome have occurred in female patients,with the exception of a patient with 47XXYkaryotype and male phenotype.' Skeletal mal-formations, especially costovertebral anomaliesand scoliosis, are the most common malforma-tions outside the central nervous system.

The cause of Aicardi syndrome is still unclear,although most authors45 agree that it isprobably an X linked dominant disorder withearly embryonic mortality in males and thateach case represents a new mutation. Theapparent ocular, cerebral, and skeletal abnor-malities imply an insult to developmentbetween the fourth and seventh week ofgestation.'Most clinical and ocular features of Aicardi

syndrome have been described in the litera-ture. Chevrie and Aicardi' have provided athorough update of the syndrome in theirreview of 38 patients examined by them and146 others. All other reports"lo describingocular involvement have been literature re-views or case reports in which fewer than eightpatients were examined by the same ophthal-mologist. Such examinations by different oph-thalmologists introduce interobserver variabil-ity in the description of ocular findings.Furthermore, although the ocular features ofthe syndrome have been described, visual acu-ity has not been documented by standard tests,nor has any correlation between the severity ofocular findings and functional changes beenevaluated.The most characteristic of these ocular find-

ings is the chorioretinal lacunae, which do notchange in size and number over time.2 They aremultiple whitish or yellowish, non-elevatedlesions with variable pigmentation at their bor-ders. Although they usually cluster around theposterior pole, they can be present in theperiphery, and their size varies from 0.1 tomore than 3 disc diameters.69

Optic disc anomalies, particularly colobo-mas, are common; overall, some anomaly isreported in approximately 50% of patients.29Other ocular abnormalities described in pa-tients with Aicardi syndrome include microph-thalmos, optic nerve hypoplasia, pupil mem-brane remnants, iris colobomas, choroidalcolobomas, retinal detachment, and abnormalfoveal development.

Visual evoked potential (VEP) and elec-troretinography (ERG) responses for patientswith Aicardi syndrome have been reportedinfrequently. In one study, flash VEP elicitednormal responses in five patients and noresponse in one other patient." ERG responseshave been reported as normal in at least eightpatients,2'2 decreased in amplitude in one,"and extinguished in a microphthalmic eye ofanother patient.'0

Department ofOphthalmology, TheHospital for SickChildren, Toronto,CanadaA V MenezesT L LewisJ R Buncic

Correspondence to:Allison VM Menezes, MD,Kantor Eye Institute, 21 1 1Bee Ridge Road, Sarasota,FL 34239, USA.

Accepted for publication24 May 1996

805

on March 22, 2021 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.80.9.805 on 1 S

eptember 1996. D

ownloaded from

http://bjo.bmj.com/

Menezes, Lewis, Buncic

The few available reports that describe theprognosis for the cognitive and motor skills ofpatients with Aicardi syndrome indicate thatthe prognosis is significantly limited. Severemental retardation is common2 4; some lan-guage function has been documented in6%-14% of patients.45 Similarly, only 6%-15% of patients are able to walk.45 On theother hand, available qualitative informationabout their visual acuity suggests that the visualprognosis is better. Case reports69 of eightpatients showed that half of the patients couldfixate on a bright object and follow a brightlight. However, of the four patients who couldnot fix and follow, only one had a specificocular feature (bilateral microphthalmos) thatcould account for significant visual impair-ment.

This study was undertaken to documentvisual function and acuity in patients withAicardi syndrome, and to determine whetherthere is any relation between ocular features ofthe syndrome exhibited at birth and later visualfunction. Furthermore, we hypothesised thatlarger lacunae may be associated with poorerclinical outcome and therefore investigated therelation between these two variables.

Patients and methodsOur study sample consisted of all patients withAicardi syndrome seen at a local hospitalbetween 1975 and 1992. All of these 14 girlshad the diagnostic criteria for Aicardi syn-drome: agenesis of the corpus callosum, infan-tile seizures, and pathognomonic chorioretinallacunae. The extent of corpus callosum agen-esis was assessed by pneumoencephalography,computed tomography scan, or magnetic reso-nance imaging depending on the best imagingprocedure that was available at the time ofdiagnosis. All ocular examinations were per-formed by the third author. Thirteen of thepatients were examined at less than 1 year ofage, and their anterior segment examination,cycloplegic refraction, and funduscopicexamination were documented. The remainingpatient (patient K, see Table 1) had a seizuredisorder and was not diagnosed with Aicardisyndrome until her ophthalmic examination at11 years of age. Photographs of the funduswere also obtained for 13 of the 14 girls.We reviewed all fundus photographs and

charts in 1992 and documented the ocularfindings. Ocular features at birth and the statusof the corpus callosum remain constantthroughout life in patients with Aicardi syn-drome.2 Lacunar size, estimated in disc diam-eters, was compared with the size of a clinicallynormal optic disc. The presence or absence oflacunae at the fovea was noted. In addition, thedegree of foveal development was graded asgood if a normal foveal pit with a healthy nervefibre layer was present and poor if thesefeatures were absent or if a macular lacunainvolved the fovea. Optic disc abnormalitieswere described as hypoplastic if the discappeared smaller than normal, as coloboma-tous if a large excavation of the disc waspresent, and as dysmorphic if any otheranatomical abnormality was present. Discs

with only mildly anomalous vessels were classi-fied as normal. Microphthalmos was definedclinically as a small eye with corneal diameterless that 10 mm.

Flash VEP and ERG had been documentedin five patients between the ages of 8 monthsand 7 years. ERG responses were recordedbinocularly in three girls using a single red flashwith skin electrodes. Bright flash VEP res-ponses were also documented binocularly inthree girls.We contacted the families of all nine patients

who were still alive in 1992 to determine thegirls' level of function and to offer the familiesthe opportunity to have their children's currentvisual acuity documented. We were able todocument visual acuity quantitatively in onlysix of these patients because they currently liveall across Canada. Although it would have beenoptimal to use the same method of evaluatingvisual acuity in all patients, this was not possi-ble. Instead, we used the most rigorous visualacuity test available, depending on the limita-tions of patients' cognitive and motor skills. Weused the standard Snellen method for each eyeof one patient, the Sheridan-Gardner singleletter test for each eye of another patient,pattern VEPs in each eye of two patients, andpreferential looking techniques for five eyes ofthree patients. We used parental reports orclinical assessment to document visual func-tion qualitatively in the remaining threepatients who were still alive.

Pattern VEPs and preferential looking arenow established methods for estimating visualacuity in non-verbal subjects. Details of theprocedures we used for each test are providedelsewhere.4 15 Each of these tests was per-formed by teams of trained observers at thehospital. Briefly, acuity was assessed withpattern VEPs by recording monocular res-ponses to pattern reversing checkerboardsand determining the smallest checks thatelicited a response different from that elicitedby a plain grey field. Acuity was assessed withpreferential looking by determining the nar-rowest stripes that the child systematicallychose over a plain grey stimulus 75% of thetime.

Visual function for seven of the nine patientswho were still alive in 1992 was compared withthe clinical appearance of the macula andfovea. Patients H and J were excluded since thestate of their foveae was unknown. We alsoexcluded the right eye of patient N because oftotal retinal detachment. For the remaining 13eyes, we used a three point scale to assign avisual function score and to assign a macularappearance score. Visual function was gradedas 1 if acuity was 2 or more octaves (an octaveis a halving or a doubling of a value) belownormal for a given test at a given age, or if thechild was visually inattentive, as observed bythe parents. Grade 2 indicated visual functionthat was 1 octave below normal. Grade 3described a child whose visual acuity was nor-mal or low normal, or who was able to fixate onand follow an object, as observed by parents orfound by clinical examination. For example,the visual function of a 9-month-old patient

806



j.com/

Br J O


eptember 1996. D

ownloaded from

http://bjo.bmj.com/

Role of ocular involvement in the prediction of visual development and clinical prognosis in Aicardi syndrome

Table 1 Clinicalfeatures of 14 patients with Aicardi syndrome

Outcome (1992 or

Ocularfeatures at birth Abnormalities Visualffunction before death)*

Lacunae Good Largestinvolve foveal lacuna Age

Patient Eye maculat development (DD) Optic disc Other ocular Age tested Result (years) Score

Non-survivors:A R - 0.5-1 H, D 1 1

L + + 0.5-1 C, DB R + (F)A - > 5 C C (retinal), 2 1

MicrophL + (F) - > 5 Pigmented

C R + (F) - 1-5 Pigmented 3 1L - 1-5 Pigmented Microph, R/D

D R + - > 5 D 13 1L - + 1-5 H,D

E R + (F) - > 5 D Microph, XT 14 1L - - 1-5 H,D

Survivors:F R - - 0.5-1 H, D 9 months 24 arc min§ 2 1

L + (F) - 1.5 H, D 9 months 24 arc min§Both 8 months VEP (N), ERG (N)

G R - + 1-5 Drusen 1 year 7.5 arcmin (N) 3 32 years 5 arc min (N)§

L - + None seen Drusen 1 year 7.5 arc min (N)2 years 4 arc min(N)§

H R - 0.5-1 C 4 years Fix and follow 4 2L - 1-5 N Binocularly$

I R - + None seen H Post embryotoxon 3 years 20/30 (N)** 5 4L - + 0.5-1 D 20/30 (N)**Both 1 year VEP (N)

J R - 1-5 Pigmented 8 years Inattentive 8 1L - None seen Pigmented 8 years InattentiveBoth 3 years VEP, ERG (N)

K R - + < 0.5 N Anisometropia 11 years 201400ft 12 4L - + 0.5-1 Anom 20/25(N)ff

vesselsL R - - > 5 H, D 7 years 15 arc min 14 1

L - - 0.5-1 H 30 arc minBoth ERG (N)

M R - + > 0.5-1 H 16 years Fix and follow$$ 16 1L - + > 5 D Fix and follow*t

N R > 5 D Microph, R/D 19 years Untestable§ 20 1L - + > 5 H, D 1.5 arc min (N)§

H = hypoplastic; D = dysplastic; C = colobomatous; Microph = microphthalmos; R/D = retinal detachment; XT = exotropia; dvpt = development; VEP = visualevoked potential; N = normal; ERG = electroretinography response; Post = posterior; Anom = anomalous.* Outcome score: 1 = immobile, no language; 2 = rolls, speaks some words; 3 = crawls, communicates with words; 4 = walks, speaks sentences.t + = yes; - = no; (F) = and fovea.* Lacunae also involve fovea.§ Preferential looking acuity.Pattern visual evoked potential acuity.¶ Estimation of visual function by parental observation.** Sheridan-Gardner visual acuity test.tt Snellen visual acuity.t$ Estimation of visual function by clinical examination.

who obtained a value of 24 arc minutes on tests

of preferential looking acuity would be grade 2because 24 arc minutes is approximately 1octave below the normal limit of 13 arc

minutes for this test and this age group.'5 Simi-larly, the visual function of a 1-year-old patientwho obtained a value of 7.5 arc minutes on

tests of pattern VEP acuity would be graded as

3 because 7.5 arc minutes is within normallimits for this test and this age group.'5

Macular appearance was classified as grade1 if the fovea was involved by lacunae and thuspoorly developed. Grade 2 indicated eitherpoor foveal development or macular, but not

foveal, lacunae. Grade 3 indicated good fovealdevelopment with no macular lacunae.

Clinical outcome measures were defined bymobility and language function and docu-mented up to 1992 or before death. We chosethe degree of mobility and language function as

measures of clinical outcome on the basis of a

study by Eyman et al."6 In a review of over 7000individuals who had severe deficits in cognitivefunction, Eyman and associates showed that

severe mental retardation is associated withdecreased life expectancy, particularly for thosewho are immobile.

Since mobility corresponded perfectly withlanguage skills, clinical outcome was classifiedon a four point scale: 1 represented immobilityand no language; 2 the ability to roll and tospeak some intelligible words; 3 the ability tocrawl and communicate with words; 4 the abil-ity to walk and speak in sentences. The Spear-man rank order correlation was used to evalu-ate the relation between visual function andmacular appearance. The probability of astatistically significant relation was assessed bydetermining the p value of the Spearman rankorder correlation coefficient (rho). A similarrank order correlation test was performed toevaluate the relation between the size of thelargest chorioretinal lacuna in a given patientand that patient's clinical outcome score.

ResultsTable 1 describes the clinical findings for all 14patients. All patients had chorioretinal lacunae

807



j.com/

Br J O


eptember 1996. D

ownloaded from

http://bjo.bmj.com/


0.5

1

2

4

8

16 _

32

o * PL acuityA * Pattern VEPO*Sheridan-Gardinero m Snellen

F

oN

L A

0.5 1 2 4 8 16Age (years)

Figure1 Visual acuity as afunction of age at test in six patients with Aicardi syncThe right vertical axis shows the denominator (Y) of the Snellen fraction in feet. Eacgives the acuity for one eye as measured by preferential looking (PL), pattern visualpotentials (VEP), the Sheridan-Gardiner single letter test, or the Snellen test. Filledrepresent data for the right eye and open symbols, data for the left eye. Letters adjaceieach symbol refer to the patients described more fully in Table 1. The shaded area shcrange of values obtainedfrom normal children tested monocularly with preferential k(see Lewis and Maurer'). For pattern VEPs, values of at least 7.5 arc minutes are tnormal limits after 6 months of age."

that were characteristic of Aicardi syndlacunar size ranged from less than 0.!diameters to greater than 5 disc diariThey often clustered around the pospole, although large lacunae were someseen in the mid periphery. Monocular lawere present in three (21%) of the 14 giwhom two (patients G and I) had goodaculity in each eye and one (patient J) wasally inattentive with abnormal flash VEPLTSome optic disc abnormality was fou

26 of 28 eyes. Eight (57%) girls had dyphic discs, three (21%) had colobomas,were unilateral in this series, and eighthad mild or moderate optic nerve hypolNone of the optic discs was severely hypitic. No simple optic nerve atrophy was f

Figure 2 Photograph of thefundus of the left eye ofpatientN showing typicalchorioretinal lacunae that spare the macula.

Severe ocular abnormalities were present infour patients (patients B, C, E, and N). Thesepatients were also immobile. These four girls

10 had unilateral microphthalmos; the retina intwo of these eyes was detached. One girl

20 (patient N) had good visual acuity with theother eye. The other three girls did not survive.

40 $> Although no information about their visual4

function is available, their foveae were all

80 poorly developed.0 Visual function in the nine patients who sur-X vived is also described in Table 1. For six of

160 = these nine patients (that is, patients F, G, I, K,.J L, N), we obtained quantitative estimates of

320 visual acuity and their results are illustrated inFigure 1. Visual acuity was in the normal to

640 low normal range in both eyes of patient G,whether tested with preferential looking or pat-tern VEP, and in both eyes of patient I. Visualacuity was also normal in one eye of each of

irome. two additional patients (patients K and N).ch point None of these six eyes demonstrated lacunaeevoked involving the fovea; all had a normal appearingsmosfovea (Fig 2).

mws the Three of the six patients (patients F, K, andgoking L) had visual acuity results below normal in atwithin least one eye. Right anisometropic amblyopia

was probably the basis of the poor visual acuityIrome: in the right eye of patient K.5 disc For the remaining four of the nine patientsieters. who were still alive in 1992, visual function wassterior estimated qualitatively from parental reports or,times clinical examination. Two of these patientsLcunae (patients H and M) were able to fixate on andirls, of follow an object. The foveae of patientM werevisual normal, but the state of foveal development ofs visu- patient H could not be determined from avail-s. able photographs of the fundus. Patient J wasmd in visually inattentive. Although her chart reviewIsmor- indicated that the maculae were uninvolvedwhich with lacunae, the development of her foveae(57%) could not be adequately determined from pho-plasia. tographs of the fundus. Finally, the right eye ofoplas- patient N was assumed to have no visual func-Found. tion: the retina was totally detached and

preferential looking acuity tests could not beused because the eye remained closed whenthe left eye was patched.

Figure 3 summarises the relation betweenthe visual function score and the macularappearance score for the 13 eyes that could beclassified on both measures. As shown inFigure 3, all but three eyes (the right eye ofpatient F, the right eye of patient K, and theleft eye of patient L) showed a perfectcorrespondence between the visual functionscore and the macular appearance score. ASpearman rank order correlation for all 13 eyesconfirmed a significant relation betweenthese two variables (rho=0.72, p<0.01), indi-cating that good foveal development is associ-ated with good visual function in Aicardisyndrome.

Flash ERG or flash VEP responses weremeasured in four of the nine surviving patients.Flash ERGs were normal in all three patientstested. However, all three had abnormal visualfunction as assessed by preferential looking(patient F), pattern VEPs (patient L), orparental observation (patient J).

c

EC._

-

._

808



j.com/

Br J O


eptember 1996. D

ownloaded from

http://bjo.bmj.com/


3

0)

oC.)U'a0

C

Culi

G,l,M 2.G,I,K,M,N Ooc

F02 _

1F0

L0

L0

1 2Macular score

0c)N.)

Cuc

3K ,* 0

CD-J

3J3

Figure 3 The relation between visualfunction andmacular appearance in 13 eyes of seven patients that couldbe classified on both measures. Higher values representbetter scores (see textfor details). Filled circles represent theright eye and open circles represent the left eye. Lettersidentify the patients listed in Table 1.

Flash VEPs were abnormal in one patient(patient J), a finding that was consistent withparental reports of visual inattentiveness.Moreover, flash VEPs were normal in a secondpatient (patient I) who had normal acuitymeasured on the Sheridan-Gardner test. How-ever, they were also normal in the third patient(patient F) who had abnormal preferentiallooking acuity.A poor clinical outcome was evident in 10

(71%) of 14 patients (Table 1): these 10 were

immobile and had no language skills (that is, a

clinical outcome score of 1). All five patientswho died fell into this category before death.Of the nine patients who were alive, five had a

poor outcome score of 1 (patients F, J, L, M,and N). Good visual function, evaluated eithersubjectively by parents or objectively, was

found in all four of the girls who could ambu-late. However, two ofthose who were immobile(patients M and N) also had good visual func-tion.

Figure 4 shows the relation between clinicaloutcome score and the size of the largestlacuna at birth. The six patients (patients B, D,E, L, M, and N) with the largest lacunae (5 discdiameters) also had the worst clinical outcomescore. Moreover, the two patients (patients I

and K) with the best clinical outcome scoresalso had the smallest lacunae. A Spearmanrank order correlation for all 14 patientsrevealed a moderate but significant correlationbetween lacunar size and outcome score

(rho=0.63, p<0.02).We also considered the association between

the neuroradiological findings at time ofdiagnosis ofAicardi syndrome and the ultimateclinical outcome. All our patients demon-strated partial (patients G, K, and M) or totalagenesis of the corpus callosum either on

pneumoencephalography (patients E, D, andM), computed tomography scanning (patientsA, B, C, F, G, H, I, J, L, and N), or magneticresonance imaging (patient K). Although allthree of the girls with partial agenesis had goodvisual function, two were able to ambulate and

>5 r

1-5 H

0.5-1K

4 B,D,E,L,M,N

C,F,J H G0co, 0 0

A0

IK&P

41 2 3Clinical outcome score

Figure 4 The relation between the largest lacunar size andclinical outcomefor all 14 patients in our series. Each pointrepresents the results for one patient. Higher values on the Xaxis represent a better clinical outcome score (see textfordetails). Letters adjacent to each point identify the patientslisted in Table 1.

communicate; one was not. Other neuroradio-logical abnormalities included porencephaliccysts (patients D, G, I, and J), decreased corti-cal tissue (patients B, C, D, F, M, and N), andcortical heterotopias (patients B, D, and I). Ofthe three girls with heterotopias, two have died(at age 13 years and 2 years); one can walk andspeak, and has normal visual acuity onSheridan-Gardner single letter testing.

DiscussionWe have described the ocular findings, visualfunction, and clinical outcome, as measured bymobility and language, in a series of 14 patientswith Aicardi syndrome. We found that thedegree of ocular involvement at birth can sug-gest the child's visual, language, and ambula-tory potential. Specifically, good visual func-tion can be expected if the chorioretinallacunae do not involve the fovea and if normalfoveal development is apparent on clinicalexamination. Children with smaller chorioreti-nal lacunae at birth appear to have a betterultimate clinical prognosis: they have bettermobility and language skills. Since the locationand size of the chorioretinal lacunae do notchange during the patient's lifetime, theirstatus can be evaluated at any time to assist indetermining the patient's visual and clinicalprognosis.

Visual development can sometimes be esti-mated by flash ERGs or flash VEPs. Otherauthors2"'-" have found normal responses inmost patients tested with flash VEP or ERG.They did not, however, make any attempt tocorrelate these findings with more rigorousmethods of testing visual acuity. In our series,neither flash VEP nor flash ERG recordedbefore visual acuity testing was useful inpredicting ultimate visual function in patientswith Aicardi syndrome.Good visual fumction appears to be pre-

served, as expected, if the chorioretinal lacunaedo not involve the fovea and if the foveal devel-opment appears normal on funduscopicexamination. Previous histopathological stud-

809



j.com/

Br J O


eptember 1996. D

ownloaded from

http://bjo.bmj.com/


ies7 of the chorioretinal lacunae show hypopig-mentation and abnormal differentiation of theretinal pigment epithelium, as well as adisorganised sensory retina. Thus, poor visualfunction should be anticipated when lacunaeinvolve the fovea, and good visual functionwhen the lacunae do not involve a welldifferentiated fovea. Our findings of a signifi-cant correlation between visual function andmacular appearance confirmed this prediction.Although good visual outcome may be

suggested by a normal appearing fovea devoidof lacunae, other factors may play an importantrole in visual development. Ocular abnormali-ties such as retinal detachment and ani-sometropia contributed to poor vision in ourpatients. Even if the anterior visual pathwaysappear intact, the posterior pathways, whichare important in processing and interpretingvisual signals, may be abnormal. These poste-rior pathways may be affected by sedation sec-ondary to anticonvulsants, seizure activity, orneuroanatomical abnormalities. Neuroradio-logical studies248 of patients with Aicardisyndrome have illustrated such abnormalitiesas partial or total agenesis of the corpus callo-sum, cortical heterotopias (resulting fromabnormal neuronal migration), ventriculardilatation, porencephaly, posterior fossa cysts,and cerebellar hypoplasia. In addition, abnor-malities in cortical cytoarchitecture, malforma-tions of cerebral gyri, and agenesis of thecorpus callosum have been confirmedhistopathologically. .17-19 Despite the evidenceof large brain lesions, however, no opticatrophy that could have resulted from suchabnormalities was identified.

Partial agenesis of the corpus callosum maybe associated with good vision, since all threeof our patients had good visual function. How-ever, partial agenesis of the corpus callosumdoes not necessarily predict good clinicaloutcome: only two of these three patients weremobile. Cortical heterotopias, surprisingly, didnot predict either visual or clinical outcome. Ofthe three patients with heterotopias, two wereimmobile and had no language before death,but one had normal visual acuity and was ableto walk and speak. Note, however, that had oursample been larger, we might have found a sig-nificant relation between the status of the cen-tral nervous system and clinical outcome.

Since patients with Aicardi syndrome areoften developmentally delayed, clinical meas-urements of visual acuity may also be affectedby their mental retardation and lack ofcooperation. On ophthalmic examination, de-velopmentally delayed children who are overthe age of 30 months and have normal eyeshave been shown to have below normal prefer-ential looking grating acuities (Wyngaardenand Lewis, unpublished data). Under that age,however, Wyngaarden and colleagues (unpub-lished data) found no relation between cogni-tive level and grating acuity. This suggests thatour 19-year-old patient N, whose gratingacuity in her left eye was in the low normalrange, might have visual acuity well within nor-mal limits if her marked developmental delaywere not present. In contrast, there would be

no reason to expect that developmental delaycontributed to the poor grating acuity ofpatient F, tested at 9 months of age. Nor wouldthere be any reason to believe that develop-mental delay affected the objective patternVEP results.Good visual function does not consistently

predict a better clinical outcome. Four girlswho had good vision in at least one eye all hadsome ambulatory and language skills. How-ever, two patients who were immobile and hadno language ability still demonstrated goodvisual function in at least one eye.When we evaluated the size, location, and

symmetry of the lacunae to determine whetherthese factors could predict eventual clinicaloutcome, we found that the size of the largestchorioretinal lacuna was moderately but sig-nificantly correlated with outcome measures ofmobility and language skills. All patients whoselacunae were larger than 5 disc diameters in atleast one eye were immobile. Furthermore, ofthe four patients who had a good clinicaloutcome, one was only mildly affected in botheyes, and the other three had asymmetric ocu-lar involvement with small or no lacunae in onegood eye. Despite the potential inaccuracy ofsizing the lacunae by gross visual comparisonwith a clinically normal optic disc, we were stillable to show a significant correlation withmobility and language outcome. The size ofthe largest lacuna did not, however, appear tocorrelate with visual function. Rather, visualfunction correlated with the degree of fovealdifferentiation or involvement by the lacunae.The chorioretinal lacunae can be asymmet-

ric in size, as in two of our patients who hadlacunae in one eye that were larger than 5 discdiameters and in the other eye, smaller than 1disc diameter. Lacunae can also occur in onlyone eye: monocular lacunae have been de-scribed in 8% to 14% of patients in otherreports.28 We found them in three (21%) ofour patients: two had good visual, ambulatory,and language development; one did not.Evidently, the presence of only monocularlacunae does not consistently predict clinicalprognosis.

Severe eye disease was also asymmetric inour series. Microphthalmos, found in four(29%) of our patients, was associated with reti-nal detachment in two and with a large retinaland iris coloboma in another. Microphthalmoshas been reported in 21%-44% of patients.248Although it is usually unilateral, it is difficult todetermine from the available reviews whetherthe other eye was also severely involved. In ourseries, the other eye of the four patients withmicrophthalmos was of normal size with noretinal detachment. In fact, the normal sizedeye of patient N had normal visual acuity, asmeasured by preferential looking. Three of thefour patients with such severe eye disease died,one by age 14 years and two by age 3 years.The remaining girl who survived is immobileand has no language skills. Clearly, severe ocu-lar involvement is often unilateral and appearsto be an indicator of poor prognosis.We found some optic nerve abnormality in at

least one eye of all of our patients, although

810



j.com/

Br J O


eptember 1996. D

ownloaded from

http://bjo.bmj.com/


others have found a much lower frequency(50% ofpatients)." This discrepancy probablyreflects our inclusion criteria, since we, unlikeothers, classified optic nerve drusen and mildlyhypoplastic, dysplastic, or pigmented discs as

abnormal. Optic nerve abnormalities, however,do not seem to be associated with eitherimpaired vision or poor clinical outcome.Despite the presence of dysmorphic, hypoplas-tic, or colobomatous discs in four of our

patients, they all demonstrated good visualfunction as long as the foveal development inthe affected eye was good. Of the nine patientsstill alive, six had significant disc abnormalities,a finding that illustrates the poor prognosticvalue of this feature.

In conclusion, the description and frequencyof the ocular features in our series of patientswith Aicardi syndrome are similar to those dis-cussed by other authors, with the exception ofthe frequency of optic disc abnormalities. Toour knowledge, however, this is the firstsystematic evaluation of visual acuity in pa-tients with Aicardi syndrome and of theprognostic relation between ocular features atbirth and clinical and visual outcome.Our findings emphasise the role of the oph-

thalmologist both in the diagnosis and man-

agement of patients with Aicardi syndrome.The diagnosis can be confirmed from the iden-tification ofpathognomonic chorioretinal lacu-nae. Visual acuity can be documented withstandard clinical tests such as the Snellen andSheridan-Gardner tests, or with pattern VEPand preferential looking, if available. Unfortu-nately, in our series, the role of ERGs andVEPs with a flash stimulus had limited predic-tive value for estimating visual function. Goodvisual function can be anticipated if the foveaof at least one eye appears normally developedand is uninvolved by lacunae. The child shouldreceive regular ophthalmic care and be moni-tored for anisometropia, refractive error, andretinal detachment. Severe involvement of oneeye with microphthalmos may indicate a poor

prognosis. All patients should receive maximaltherapeutic assistance to enable them todevelop to their full potential since this cannotbe predicted with certainty in infancy. How-ever, the presence of predominantly small cho-rioretinal lacunae, especially if one eye isnormal or only mildly involved, suggests a bet-

ter prognosis for mobility and language devel-opment in children with Aicardi syndrome.

This paper was prepared with the assistance of EditorialServices, The Hospital for Sick Children, Toronto, Ontario,Canada.

Presented in part as a poster at the Canadian Congress ofNeurosciences, Toronto, Ontario, Canada, June 1993.

Supported in part by National Institutes ofHealth (Bethesda,MD) Grant EY-03475.

1 Aicardi J, Lefebvre J, Lerique-Koechlin A. A new syndrome:spasms in flexion, callosal agenesis, ocular abnormalities.Electroencephalogr Clin Neurophysiol 1965;19:609-10.

2 Chevrie JJ, Aicardi J. The Aicardi syndrome. In: Pedley TA,Meldrum BS, eds. Recent advances in epilepsy, vol 3.Edinburgh: Churchill Livingstone, 1986:189-210.

3 Hopkins UJ, Humphrey I, Keith CG, Susman M, Webb GC,Turner EK. The Aicardi syndrome in a 47, XXY male. AustPaediatry 1979;15:278-80.

4 Donnenfeld AE, Packer RJ, Zackai EH, Chee CM, SellingerB, Emanuel BS. Clinical, cytogenetic, and pedigreefindings in 18 cases of Aicardi syndrome. Am J Med Genet1989;32:461-7.

5 Neidich JA, Nussbaum RL, Packer RJ. Heterogeneity ofclinical severity and molecular lesions in Aicardi syndrome.J Pediatr 1990;116:911-7.

6 Hoyt CS, Bilison F, Ouvrier R, Wise G. Ocular features ofAicardi's syndrome. Arch Ophthalmol 1978;96:291-5.

7 Del Pero RA, Mets MB, Tripathi RC, Torczynski E.Anomalies of retinal architecture in Aicardi syndrome.Arch Ophthalmol 1986;104:1659-64.

8 Bertoni JM, von Loh S, Allen RJ. The Aicardi syndrome:report of 4 cases and review of the literature. Ann Neurol1979;5:475-82.

9 Denslow GT, Robb RM. Aicardi's syndrome: a report offour cases and review of the literature. Y Pediatr OphthalmolStrabismus 1979;16: 10-5.

10 Rousselie F, Fontaine M. Anomalies chorio-retiniennes.Spasmes en flexion. Agenesie calleuse. Arch Opht (Paris)1979;31(6-7):539-46.

11 Yamamoto N, Watanabe K, Negoro T, Matsumoto A, HaraK, Miyazaki S, et al. Aicardi syndrome. Report of 6 casesand a review ofJapanese literature. Brain Dev 1985;7:443-9.

12 Weleber RG, Lovrien EW, Isom JB. Aicardi's syndrome-case report, clinical features and electrophysiologic studies.Arch Ophthalmol 1978;96:285-90.

13 Gerhard JP, Juif JG, Luckel JC, Mack M. A propos d'unenouvelle observation de syndrome de spasme en flexion,agenesie calleuse et anomalies chorio-retiniennes (syn-drome Aicardi). Bull Soc Ophtalmol Fr 1973;3:363-6.

14 McCulloch DL, Skarf B. Development of the human visualsystem: monocular and binocular pattern VEP latency.Invest Ophthalmol Vis Sci 1991;32:2372-81.

15 Lewis TL, Maurer D. Preferential looking as a measure ofvisual resolution in infants and toddlers: a comparison ofpsychological methods. Child Dev 1986;57:1062-75. (Ex-perimnent 3)

16 Eyman RK, Grossman HJ, Chaney RH, Call TL. The lifeexpectancy of profoundly handicapped people with mentalretardation. N EnglJ Med 1990;323:584-9.

17 deJong JGY, Delleman JW, Houben M, Manschot WA, deMinjer A, Moi J, et al Agenesis of the corpus callosum,infantile spasms, ocular anomalies (Aicardi's syndrome)-clinical and pathologic findings. Neurology 1976;26: 1152-8.

18 Font RL, Marines HM, Cartwright J, Bauserman SC.Aicardi syndrome-a clinicopathologic case report includ-ing electron microscopic observations. Ophthalmology1991;98:1727-31.

19 McMahon RG, Bell RA, Moore GRW, Ludwin SK.Aicardi's syndrome. A clinicopathologic study. ArchOphthalmol 1984;102:250-3.

811



j.com/

Br J O


eptember 1996. D

ownloaded from

http://bjo.bmj.com/

role in andclinical aicardi - british journal of ophthalmology · aicardi syndrome seen at a local...

Documents